Kaiser Permanente Northern California (KPNC) and its Division of Research (DOR) propose a set of three evaluations of KPNC system-level efforts to prevent development of diabetes in high-risk members. One or more of these studies would be well-suited for pooling with data from other systems in this research network on similar interventions. The applicant research team is highly experienced in working with KPNC clinical and operational leaders to evaluate natural experiments in diabetes care, including recent evaluations as part of the CDC/NIDDK-sponsored TRIAD study. We believe this is an is an opportune moment to examine whether the population approaches that have been effective for patients with diabetes can be adapted to support high-risk pre-diabetic patients to modify lifestyles and prevent progression to diabetes. The primary evaluation is of a KP - employer partnership that has adapted its worksite wellness program to risk stratify all employees using electronic health record (EHR) data and to use intensive outreach, incentives, and coaching (web, telephone, or in-person) to support lifestyle change in high-risk employees. Intervention and control employer sites will be selected in consultation with CalPERS, a large benefits management group representing multiple employers. We expect to identify approximately 750 high-risk employees at each site. Impact will be assessed using both EHR and patient surveys (via mail, phone, or internet at baseline and 12 months) to measure changes in body mass index (BMI), physical activity, blood pressure and lipid levels. The first ancillary study evaluates a telephonic wellness coaching program established in January 2010. It is intended to be part of a portfolio of coaching support that also includes online and in-person services. It is operated by KPNC's Regional Health Education Department (RHE), and will target as many as 10,000 high risk persons per year when fully implemented. We will work with RHE to evaluate program impact among high risk members identified either from EHR or health risk appraisal data, or from a combination. Surveys at baseline and 12 months will again measure changes in BMI, physical activity, blood pressure and lipids. We will also assess readiness to change, preferences for type of coaching, and correlates of these preferences. A second ancillary study examines a regionwide program implemented in 2006 that aims to ensure postpartum screening for diabetes, using a 75-gram oral glucose tolerance test, in all women with gestational diabetes mellitus (GDM). The program conducts outreach to those who prove to have pre-diabetes at testing. The program will be evaluated by examining changes in population incidence (over 5 years postpartum) in cohorts of all women with GDM from 2001-05 vs 2006-2010 (n=12,000 in each cohort). Incidence will be divided into less severe (often screen-detected) and more severe (with A1c >8% or requiring two medications). We hypothesize that the program will lead to decreases in more severe diabetes but to increases in milder disease. We also list a number of natural experiments in diabetes care that we would be able to evaluate.